The ultimate goal of this proposal is to utilize a neurobotics paradigm to assist trunk and limb controls by applying force at the pelvis during locomotion in normal and injured rats. We will also use more standard physiology. Together these two approaches provide tools to examine normal and post- injury corticospinal organization, and the control of trunk and hind-limbs. We seek to understand and improve trunk control after SCI, and to examine its development, modularity and its plasticity in intact and spinal cord injured (SCI) rats. We have three Specific Aims: Aim 1 : We will identify physiological and biomechanical differences in the use of trunk and leg muscles and the associated motor cortical activity between (1) adult rats with neonatal spinal transections with weight support and (2) adult rats with neonatal spinal transections without weight support and (3) normal rats. Aim 2: We will examine how normal rats alter neural and motor activity in response to neurorobotic interventions which generate lumbar actions. We will test (1) robot elastic force-field actions that are extrinsic or intrinsic but not contingent on neural activity, and (2) force-field actions directly contingent on features of neural activity (neurorobotic control). Aim 3: We will compare how neonatal injured SCI rats with good or partial weight support alter neural and motor activity in response to neurorobotic interventions which generate lumbar actions. We will test (1) robot elastic force-field actions that are extrinsic or intrinsic but not contingent on neural activity, and (2) force-field actions directly contingent on features of neural activity (neurorobotic control). The research here can contribute to the clinical mission of providing therapies for SCI and other trauma in a range of ways. First, by furthering our understanding of cortical and spinal integration, in normal and neonatal SCI rats with and without weight support (Aim 1) we will provide information on how best to assess and optimize recovery in rat models of injury and perhaps beyond. Second, by developing an animal model of pelvis interaction rehabilitation (Aim 2 and 3), we will provide basic data on what advantages or additional benefits this framework may have in a model where more invasive recording is feasible. This may be of fairly direct relevance to pelvic assistive devices under development for the clinic. Third, if the intact neonatal injured rat or the adult injured rats can learn to use a neurorobotic control of pelvis, we will have demonstrated a neural bypass strategy for trunk and legs which may be extended to intraspinal stimulation, FES or other higher degree of freedom control methods for the musculoskeletal and spinal systems in human SCI. PUBLIC HEALTH RELEVANCE Brain Machine Interfaces and novel prosthetics will in future require controls of the trunk as well as the limbs for injuries of spinal cord causing paraplegia or tetraplegia. Currently there is no animal model of rehabilitation and neurorobotics of the trunk. Trunk is essential for coordinated locomotion and action. We develop an animal model of trunk robotic rehabilitation and brain machine interface control.